Information Architecture Testing

The information architecture in a website or application refers to the actual content and how it is organized. The ultimate goal of information architecture testing is to improve the findability of content and help users to do what they came to do at the website or app.

Why You Should Do It

One of the major frustrations users have with websites or applications is not finding the information, path, or specific functionalities that are needed to complete a task.

Content organization (e.g., navigation menus, sub-menus, points of entry, etc.) and content labeling (e.g. menu labels) are often major barriers to findability, navigation, and good user experience.

The Approach

The most common techniques to uncover information architecture issues are Tree Testing and Card Sorting.

• Tree Testing: Participants are tasked with finding different elements or pieces of information in stand-alone information structures (e.g. menus) outside the context of a visual design or other functionalities (e.g. search). This technique helps isolate navigation issues and identify which menu groups and labels may be leading users astray.
• Card Sorting: Participants are asked to group items (e.g., pages, content categories, menu labels, etc.) into content categories. They may be asked to create their own categories (open sort) or to assign them to pre-defined categories (closed sort). This technique can help create a navigation structure that’s better aligned with users’ mental models and expectations.

When Is It Appropriate to do Information Architecture Testing?

Information Architecture testing can be conducted at different stages of the design process:

• Before the development of design concepts: Testing the information architecture of the current website or application can provide valuable insights into navigation problems affecting the user experience that can be fixed with a new design.
• Early concept development: Information architecture testing in combination with usability testing of low-resolution prototypes (e.g. paper prototypes) in an iterative process can help detect problems with content organization and user errors.
• Prototype development: Before launching the “final” version of the website or app, you will benefit from testing navigation-enabled (clickable) prototypes to detect navigation problems.

These techniques are often combined with qualitative usability testing. However, they can be conducted on their own or as part of quantitative usability testing. They can be moderated or unmoderated.

Target Audience Recruitment

Relevant Insights can recruit qualified B2C and B2B participants through our sample provider partners. Relevant Insights can also recruit participants from your customer database and will implement procedures to protect any identifiable personal information.

As an alternative, you may also recruit directly from your own customer database if data privacy policies restrict customer data sharing. In these cases, we will provide the necessary data collection tools.

After discussions with you, we determine the sample parameters and assist in developing screeners. We also manage participant incentives.

Implementation

In direct collaboration with you, Relevant Insights will:

1. Develop tasks, scenarios, before-task, and follow-up questions to support the study objectives.
2. Recommend sample size, target audience, and sample sources.
3. Set up and host usability tests with appropriate tools for both moderated and unmoderated information architecture testing.
4. Rent facilities or online platforms as needed.
5. Moderate in-person or remote user test sessions.

Analytical Plan

Metrics from information architecture testing vary depending on the technique used.

Tree Testing

• Findability success rate.
• Time to find requested elements.
• Error rate.
• Perceived level of difficulty of the tasks.
• First clicks.
• Followed paths.

Card Sorting

• Frequency of group categories.
• Frequency of combinations of group categories (similarity matrix).
• Content grouping visualization (dendrograms).
• Popular grouping strategies.

Reporting

Relevant Insights can provide different levels of reporting based on budget and time constraints including:

1. Short summary report with key findings.
2. A full, detailed report with analysis and graphics in an appealing format.
3. Audio and video recordings, if included in the study set up.

Typical Project Duration

Five to eight weeks.

Factors that can affect project duration include:

• The incidence rate of the target sample: The lower the incidence rate, the longer we need to stay in the field to gather the required data.
• Client team responsiveness: A delayed response to requests for feedback at different steps of the project, will stall a project and affect the delivery date.
• The number of participants needed to be recruited.

Investment

The cost will vary depending on:

• Testing technique and approach (moderated or unmoderated) selected.
• Sample specifications and sample size.
• The number of participants.